INDEX

Explanations

modified versions or adaptations of something

oai_token-act-pair · gpt-3.5-turbo

references to items or concepts that have been altered or adapted

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

çĦ

-0.84

ä

-0.73

HI

-0.70

OPA

-0.69

Water

-0.67

vor

-0.66

riel

-0.66

tale

-0.65

gary

-0.65

True

-0.64

POSITIVE LOGITS

atile

0.88

xual

0.84

 modification

0.83

 modifications

0.82

icum

0.79

 wrench

0.78

iations

0.78

 mitigation

0.78

hap

0.77

 organisms

0.75

Activations Density 0.015%

modified versions or adaptations of something

references to items or concepts that have been altered or adapted

No Comments

No Known Activations